(F-201) Determinants of NK cell cytotoxicity sensitivity in mouse models of breast cancer dormancy.

Introduction:: Patients with estrogen receptor-positive (ER+) breast cancer are at constant risk of recurrence despite the benefits of adjuvant therapy. One explanation is the existence of disseminated tumor cells (DTCs) that are resistant to therapy, exhibit slow or no growth, and can be reactivated. However, the lack of fully immunocompetent mouse models of ER+ breast cancer dormancy has made it challenging to understand the mechanisms behind these interactions. To fill this gap, our objective is to establish an immunocompetent model of breast cancer dormancy.


Materials and Methods:: D2.0R cell line was grown in DMEM with 10% FBS and 1x glutamax and kept at a passage < 30. For label retention experiments, tumor cells were labeled with cell trace far red and cultured in 2% FBS for 7d to allow development of a quiescent, label-retaining population and a proliferative non-label retaining population which were sorted via fluorescent activated cell sorting. For bulk RNAseq, RNA was isolated from cells using Qiagen RNeasy kit according to manufacturer instructions. QC, library preparation and sequencing were performed by the Advanced Genomics Core (University of Michigan). For single cell RNAseq, cells were sorted via FACS and maintained on ice until isolation via 10x chromium 5’ sequencing was performed by the Advanced Genomics Core. Data was analyzed in R.


Results, Conclusions, and Discussions:: Previously, we identified that D2.0R can serve as an immunocompetent model of breast cancer dormancy in bone marrow, controlled by natural killer (NK) cells. D2.0R has both a quiescent (label retaining cells, LRC) and a proliferative (non-label retaining cells, non-LRC) subpopulation. We hypothesized that these might be differentially sensitive to NK cell killing. We found the quiescent population is NK resistant, while the proliferative population is NK sensitive.

To investigate the mechanisms behind this differential sensitivity to NK cells, we performed bulk RNA sequencing of LRC vs non-LRC. We found 2648 differentially expressed genes (DEGs) between the two populations (Fig 1a, b). Next, we investigated the pathways that these DEGs represent via gene set enrichment analysis. We found that pathways suppressed in LRCs were related to proliferation, while pathways enriched were related to inflammation (Fig 1c). We also investigated the expression of major histocompatibility complex (MHC) molecules (Fig 1d) and NK ligands (Fig 1e). MHC, which serves as a negative regulator of NK cytotoxicity was up in LRCs, whereas NK ligands, which activate NK cytotoxicity were up in non-LRCs. We next used RNA-seq data to identify candidate transcription factors (TF) of these phenotypes. We identified BACH1 as a significantly predicted TF for both the DEGs in LRCs and the expression of MHC (Fig 1f). We identified SOX2 as a significantly predicted TF for both the DEGs in LRCs and the expression of NK ligands (Fig 1g). We further investigated these LRC/non-LRC populations via single cell RNAseq to investigate population heterogeneity and interactions between populations. Like bulk RNA-seq data we found that differences in cell cycle could not explain all the phenotypic differences between the two populations.

In conclusion, we investigated interactions between D2.0R tumor cells and NK cells in vitro and in vivo. We made an important observation that the quiescent, mesenchymal D2.0R CSC population is resistant to NK lysis, while the epithelial, proliferative D2.0R CSC population is sensitive to NK lysis. This is the first report to investigate the sensitivity of D2.0R cells to NK cells and investigate the mechanism of sensitivity/resistance.

Acknowledgements (Optional): :

References (Optional): :